Probe sex * INR interactions

#Libraries
library(tidyverse)
library(readxl)
library(modelr)
library(sjstats)
library(car)
#Files
volumes_file <- "~/Desktop/ELS/income_TBM/data/final_data/inr_volumes_20190218.xlsx"
age_match_ids_file <- "~/Desktop/ELS/income_TBM/data/final_data/age_matched_ids.csv"

Read in data

vols <-
  read_xlsx(volumes_file, sheet = "CSX Volumes") %>% 
  rename(
    pos_rSTG_cs = Pos_10, #superior temporal gyrus
    pos_lIFG_cs = Pos_11, #inferior frontal gyrus
    pos_lLOC_cs = Pos_12, #lateral occipital gyrus
    pos_rMFG_cs = Pos_14, #middle frontal gyrus
    pos_rSFG_cs_dlPFC = Pos_15, #superior frontal gyrus
    pos_rPCG = Pos_16, #postcentral gyrus,
    pos_rSFG_cs_motor = Pos_17,
    pos_rPTR_cs = Pos_18, #posterior thalamic radiation,
    pos_rMB_cs = Pos_19, #midbrain,
    pos_rCV_cs = Pos_20, #cerebellar vermis,
    neg_FG_cs = Neg_7, #fusiform gyrus,
    neg_lThal_cs = Neg_8, #thalamus
    neg_rHipp_cs = Neg_9, #hippocampus/CGH
    neg_lHipp_cs = Neg_10, #hippocampus/CGH/fusiform gyrus,
    neg_rThal_cs = Neg_11, #thalamus
    neg_lAG_cs = Neg_12 #angular gyrus
  ) %>% 
  left_join(
    read_xlsx(volumes_file, sheet = "Long Volumes") %>% 
      rename(
        pos_rSPL_lg = Pos_4, #superior parietal lobule
        pos_rLG_lg = Pos_5, #lingual gyrus,
        pos_lITG_lg = Pos_6, #inferior temporal gyrus,
        pos_lLG_lg = Pos_7, #lingual gyrus,
        neg_rHipp_lg = Neg_7, #hippcampus/CGH,
        neg_lSLF_lg = Neg_8, #superior longitudinal fasciculus,
        neg_lSFG_lg = Neg_9, #superior frontal gyrus
        neg_rSFG_lg = Neg_10, #superior frontal gyrus
        neg_lCerebel_lg = Neg_11, #cerebellum
        neg_rPTR_lg = Neg_12 #posterior thalamic radiation
      ) %>% 
      select(-T1_ICV, -T1_Age, -White, -INR_Sex, -Sex, -INR),
    by = "Subject"
  ) %>% 
  rename(ID = Subject) %>% 
  mutate(Male = as.factor(Sex)) %>% 
  left_join(
    read_csv(age_match_ids_file) %>% 
      rename(
        ID = ELS_ID
      ) %>% 
      mutate(
        age_matched = 1
      ),
      by = "ID"
  )

-
/
                                                                                                  

-
/
                                                                                                  

Parsed with column specification:
cols(
  ELS_ID = col_double()
)

vols_match <-
  vols %>% 
  filter(age_matched == 1)

Cross-sectional simple effect regressions

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm1Ma <- lm(pos_rCV_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm1Ma)

Call:
lm(formula = pos_rCV_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.16663 -0.05848 -0.02213  0.05670  0.19117 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)   
(Intercept)     -1.5339084  1.7176767  -0.893  0.37650   
sqrt(INR)        0.4495654  0.1330512   3.379  0.00149 **
Male1            0.5666757  0.2163699   2.619  0.01190 * 
T1_Age           0.0699678  0.0567636   1.233  0.22398   
White            0.0196378  0.0303167   0.648  0.52036   
T1_ICV           0.0002547  0.0011223   0.227  0.82150   
sqrt(INR):Male1 -0.5140601  0.2048626  -2.509  0.01568 * 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1026 on 46 degrees of freedom
Multiple R-squared:  0.2274,    Adjusted R-squared:  0.1267 
F-statistic: 2.257 on 6 and 46 DF,  p-value: 0.05426

std_beta(lm1Ma)

summary(lm2M)

Call:
lm(formula = pos_rMB_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.38276 -0.08212 -0.00292  0.08828  0.30981 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)      0.4634134  0.9866559   0.470 0.639314    
sqrt(INR)        0.3313063  0.1233348   2.686 0.008101 ** 
Male1            0.5777801  0.1554662   3.716 0.000291 ***
T1_Age           0.0153109  0.0105009   1.458 0.147064    
White           -0.0400713  0.0209360  -1.914 0.057662 .  
T1_ICV          -0.0008724  0.0007268  -1.200 0.232055    
sqrt(INR):Male1 -0.5123634  0.1473558  -3.477 0.000676 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1237 on 140 degrees of freedom
Multiple R-squared:  0.1445,    Adjusted R-squared:  0.1079 
F-statistic: 3.942 on 6 and 140 DF,  p-value: 0.001131

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm2Ma <- lm(pos_rMB_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm2Ma)

Call:
lm(formula = pos_rMB_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.34515 -0.06850  0.02252  0.06030  0.27792 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)  
(Intercept)      3.999029   2.139785   1.869   0.0680 .
sqrt(INR)        0.246811   0.165748   1.489   0.1433  
Male1            0.527849   0.269542   1.958   0.0563 .
T1_Age          -0.090490   0.070713  -1.280   0.2071  
White           -0.079966   0.037767  -2.117   0.0397 *
T1_ICV          -0.002538   0.001398  -1.815   0.0761 .
sqrt(INR):Male1 -0.449680   0.255206  -1.762   0.0847 .
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1278 on 46 degrees of freedom
Multiple R-squared:  0.2784,    Adjusted R-squared:  0.1843 
F-statistic: 2.958 on 6 and 46 DF,  p-value: 0.01584

std_beta(lm2Ma)

outlierTest(lm3F) 

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm3Ma <- lm(pos_rPTR_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm3Ma)

Call:
lm(formula = pos_rPTR_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.218653 -0.070921  0.000017  0.063801  0.237211 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)   
(Intercept)     -1.6772470  2.0470563  -0.819  0.41681   
sqrt(INR)        0.4601431  0.1585650   2.902  0.00567 **
Male1            0.7491131  0.2578608   2.905  0.00563 **
T1_Age           0.0447676  0.0676486   0.662  0.51142   
White           -0.0174304  0.0361302  -0.482  0.63179   
T1_ICV           0.0005576  0.0013375   0.417  0.67873   
sqrt(INR):Male1 -0.6152283  0.2441468  -2.520  0.01527 * 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1222 on 46 degrees of freedom
Multiple R-squared:  0.3139,    Adjusted R-squared:  0.2244 
F-statistic: 3.507 on 6 and 46 DF,  p-value: 0.006116

std_beta(lm3Ma)

outlierTest(lm4F) 

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm4Ma <- lm(pos_rSFG_cs_motor ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm4Ma)

Call:
lm(formula = pos_rSFG_cs_motor ~ sqrt(INR) * Male + T1_Age + 
    White + T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.34653 -0.09981  0.00572  0.09935  0.35439 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)   
(Intercept)     -1.996834   2.630089  -0.759  0.45159   
sqrt(INR)        0.419811   0.203727   2.061  0.04502 * 
Male1            0.952567   0.331303   2.875  0.00610 **
T1_Age           0.025999   0.086916   0.299  0.76619   
White           -0.045970   0.046421  -0.990  0.32721   
T1_ICV           0.000800   0.001718   0.466  0.64374   
sqrt(INR):Male1 -0.957334   0.313684  -3.052  0.00377 **
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1571 on 46 degrees of freedom
Multiple R-squared:  0.2228,    Adjusted R-squared:  0.1215 
F-statistic: 2.198 on 6 and 46 DF,  p-value: 0.06014

std_beta(lm4Ma)

outlierTest(lm5F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm5Ma <- lm(pos_rPCG ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm5Ma)

Call:
lm(formula = pos_rPCG ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, 
    data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.24765 -0.12809  0.01160  0.08905  0.36692 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)  
(Intercept)      1.5665622  2.6098063   0.600   0.5513  
sqrt(INR)        0.4621886  0.2021556   2.286   0.0269 *
Male1            0.5792700  0.3287485   1.762   0.0847 .
T1_Age          -0.1065636  0.0862456  -1.236   0.2229  
White           -0.0531768  0.0460626  -1.154   0.2543  
T1_ICV          -0.0005821  0.0017052  -0.341   0.7344  
sqrt(INR):Male1 -0.5274940  0.3112645  -1.695   0.0969 .
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1558 on 46 degrees of freedom
Multiple R-squared:  0.1929,    Adjusted R-squared:  0.08764 
F-statistic: 1.832 on 6 and 46 DF,  p-value: 0.1136

std_beta(lm5Ma)

outlierTest(lm6F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm6Ma <- lm(pos_rSFG_cs_dlPFC ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm6Ma)

Call:
lm(formula = pos_rSFG_cs_dlPFC ~ sqrt(INR) * Male + T1_Age + 
    White + T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.30245 -0.12260 -0.00053  0.13019  0.37256 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)   
(Intercept)      0.976568   2.935849   0.333  0.74092   
sqrt(INR)        0.646413   0.227411   2.842  0.00665 **
Male1            0.664792   0.369819   1.798  0.07880 . 
T1_Age           0.055335   0.097020   0.570  0.57122   
White            0.011561   0.051817   0.223  0.82444   
T1_ICV          -0.001726   0.001918  -0.900  0.37294   
sqrt(INR):Male1 -0.631426   0.350151  -1.803  0.07789 . 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1753 on 46 degrees of freedom
Multiple R-squared:  0.1795,    Adjusted R-squared:  0.07246 
F-statistic: 1.677 on 6 and 46 DF,  p-value: 0.1482

std_beta(lm6Ma)

outlierTest(lm7F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm7Ma <- lm(pos_rMFG_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm7Ma)

Call:
lm(formula = pos_rMFG_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.35292 -0.11674 -0.00352  0.11356  0.30581 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)  
(Intercept)      0.463282   2.922855   0.159   0.8748  
sqrt(INR)        0.512866   0.226404   2.265   0.0282 *
Male1            0.695032   0.368182   1.888   0.0654 .
T1_Age           0.106856   0.096591   1.106   0.2744  
White           -0.013195   0.051588  -0.256   0.7993  
T1_ICV          -0.001703   0.001910  -0.892   0.3772  
sqrt(INR):Male1 -0.651147   0.348601  -1.868   0.0682 .
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1745 on 46 degrees of freedom
Multiple R-squared:  0.1375,    Adjusted R-squared:  0.02503 
F-statistic: 1.222 on 6 and 46 DF,  p-value: 0.3123

std_beta(lm7Ma)

outlierTest(lm8M)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm8Ma <- lm(pos_lLOC_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm8Ma)

Call:
lm(formula = pos_lLOC_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.24193 -0.08647 -0.02296  0.09096  0.28634 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)  
(Intercept)     -0.150520   2.216343  -0.068    0.946  
sqrt(INR)        0.312772   0.171678   1.822    0.075 .
Male1            0.405342   0.279185   1.452    0.153  
T1_Age          -0.066246   0.073243  -0.904    0.370  
White           -0.009608   0.039118  -0.246    0.807  
T1_ICV           0.000470   0.001448   0.325    0.747  
sqrt(INR):Male1 -0.426981   0.264337  -1.615    0.113  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1323 on 46 degrees of freedom
Multiple R-squared:  0.1322,    Adjusted R-squared:  0.01897 
F-statistic: 1.168 on 6 and 46 DF,  p-value: 0.34

std_beta(lm8Ma)

contrasts(vols$Male) = c(0, 1)
# SE female
lm9F <- lm(pos_lIFG_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols)
summary(lm9F)

Call:
lm(formula = pos_lIFG_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.56310 -0.13836  0.01717  0.14774  0.40807 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)    
(Intercept)      1.811725   1.591927   1.138 0.257035    
sqrt(INR)       -0.283084   0.130205  -2.174 0.031376 *  
Male1           -0.874929   0.250265  -3.496 0.000633 ***
T1_Age          -0.013186   0.016904  -0.780 0.436678    
White            0.056685   0.033702   1.682 0.094808 .  
T1_ICV          -0.001107   0.001170  -0.947 0.345508    
sqrt(INR):Male1  0.823775   0.237209   3.473 0.000686 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1992 on 140 degrees of freedom
Multiple R-squared:  0.09682,   Adjusted R-squared:  0.05812 
F-statistic: 2.501 on 6 and 140 DF,  p-value: 0.02492

std_beta(lm9F)

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm9Ma <- lm(pos_lIFG_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm9Ma)

Call:
lm(formula = pos_lIFG_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
    Min      1Q  Median      3Q     Max 
-0.3695 -0.1374 -0.0204  0.1234  0.3647 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)   
(Intercept)      3.258397   2.991726   1.089  0.28177   
sqrt(INR)        0.629328   0.231739   2.716  0.00929 **
Male1            1.122858   0.376858   2.980  0.00460 **
T1_Age           0.023964   0.098867   0.242  0.80956   
White            0.046655   0.052803   0.884  0.38152   
T1_ICV          -0.003203   0.001955  -1.638  0.10816   
sqrt(INR):Male1 -1.036832   0.356815  -2.906  0.00562 **
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1787 on 46 degrees of freedom
Multiple R-squared:  0.2405,    Adjusted R-squared:  0.1414 
F-statistic: 2.428 on 6 and 46 DF,  p-value: 0.04021

std_beta(lm9Ma)

outlierTest(lm10F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm10Ma <- lm(pos_rSTG_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm10Ma)

Call:
lm(formula = pos_rSTG_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.36278 -0.10451  0.00383  0.10380  0.37672 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)   
(Intercept)      0.0478521  2.9143570   0.016   0.9870   
sqrt(INR)        0.6657509  0.2257461   2.949   0.0050 **
Male1            0.8892313  0.3671117   2.422   0.0194 * 
T1_Age          -0.1136426  0.0963101  -1.180   0.2441   
White           -0.0878571  0.0514379  -1.708   0.0944 . 
T1_ICV           0.0003022  0.0019042   0.159   0.8746   
sqrt(INR):Male1 -0.7758651  0.3475874  -2.232   0.0305 * 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.174 on 46 degrees of freedom
Multiple R-squared:  0.2978,    Adjusted R-squared:  0.2063 
F-statistic: 3.252 on 6 and 46 DF,  p-value: 0.009498

std_beta(lm10Ma)

outlierTest(lm11F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm11Ma <- lm(neg_lAG_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm11Ma)

Call:
lm(formula = neg_lAG_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.29983 -0.09409 -0.01208  0.07789  0.32190 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)
(Intercept)     -0.2389498  2.3780439  -0.100    0.920
sqrt(INR)       -0.1232605  0.1842033  -0.669    0.507
Male1           -0.1439166  0.2995542  -0.480    0.633
T1_Age           0.0467741  0.0785866   0.595    0.555
White           -0.0581777  0.0419721  -1.386    0.172
T1_ICV          -0.0001192  0.0015538  -0.077    0.939
sqrt(INR):Male1  0.0465088  0.2836228   0.164    0.870

Residual standard error: 0.142 on 46 degrees of freedom
Multiple R-squared:  0.1724,    Adjusted R-squared:  0.06445 
F-statistic: 1.597 on 6 and 46 DF,  p-value: 0.1696

std_beta(lm11Ma)

outlierTest(lm12F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm12Ma <- lm(neg_lThal_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm12Ma)

Call:
lm(formula = neg_lThal_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.33125 -0.06768  0.00319  0.07834  0.19568 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)  
(Intercept)      0.3294061  1.8121853   0.182   0.8566  
sqrt(INR)       -0.2209442  0.1403719  -1.574   0.1223  
Male1           -0.4484302  0.2282749  -1.964   0.0555 .
T1_Age           0.0345056  0.0598868   0.576   0.5673  
White            0.0491573  0.0319848   1.537   0.1312  
T1_ICV          -0.0003327  0.0011841  -0.281   0.7800  
sqrt(INR):Male1  0.3707231  0.2161344   1.715   0.0930 .
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1082 on 46 degrees of freedom
Multiple R-squared:  0.2113,    Adjusted R-squared:  0.1084 
F-statistic: 2.054 on 6 and 46 DF,  p-value: 0.0774

std_beta(lm12Ma)

outlierTest(lm13F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm13Ma <- lm(neg_lHipp_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm13Ma)

Call:
lm(formula = neg_lHipp_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.228744 -0.041216 -0.004803  0.043004  0.270573 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)   
(Intercept)     -2.761818   1.821483  -1.516  0.13630   
sqrt(INR)       -0.336101   0.141092  -2.382  0.02140 * 
Male1           -0.618514   0.229446  -2.696  0.00978 **
T1_Age          -0.023982   0.060194  -0.398  0.69217   
White            0.045593   0.032149   1.418  0.16288   
T1_ICV           0.002329   0.001190   1.957  0.05648 . 
sqrt(INR):Male1  0.580188   0.217243   2.671  0.01043 * 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1088 on 46 degrees of freedom
Multiple R-squared:  0.2449,    Adjusted R-squared:  0.1464 
F-statistic: 2.487 on 6 and 46 DF,  p-value: 0.03623

std_beta(lm13Ma)

outlierTest(lm14F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm14Ma <- lm(neg_rHipp_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm14Ma)

Call:
lm(formula = neg_rHipp_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.167577 -0.061443  0.002776  0.061129  0.140605 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)   
(Intercept)     -3.9544614  1.3971300  -2.830  0.00687 **
sqrt(INR)       -0.1615196  0.1082217  -1.492  0.14240   
Male1           -0.5417792  0.1759917  -3.078  0.00350 **
T1_Age           0.0329283  0.0461706   0.713  0.47933   
White           -0.0010871  0.0246591  -0.044  0.96503   
T1_ICV           0.0026953  0.0009129   2.953  0.00495 **
sqrt(INR):Male1  0.5072121  0.1666319   3.044  0.00385 **
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.08343 on 46 degrees of freedom
Multiple R-squared:  0.2859,    Adjusted R-squared:  0.1927 
F-statistic: 3.069 on 6 and 46 DF,  p-value: 0.01305

std_beta(lm14Ma)

outlierTest(lm15F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm15Ma <- lm(neg_lThal_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm15Ma)

Call:
lm(formula = neg_lThal_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.33125 -0.06768  0.00319  0.07834  0.19568 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)  
(Intercept)      0.3294061  1.8121853   0.182   0.8566  
sqrt(INR)       -0.2209442  0.1403719  -1.574   0.1223  
Male1           -0.4484302  0.2282749  -1.964   0.0555 .
T1_Age           0.0345056  0.0598868   0.576   0.5673  
White            0.0491573  0.0319848   1.537   0.1312  
T1_ICV          -0.0003327  0.0011841  -0.281   0.7800  
sqrt(INR):Male1  0.3707231  0.2161344   1.715   0.0930 .
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1082 on 46 degrees of freedom
Multiple R-squared:  0.2113,    Adjusted R-squared:  0.1084 
F-statistic: 2.054 on 6 and 46 DF,  p-value: 0.0774

std_beta(lm15Ma)

outlierTest(lm16F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm16Ma <- lm(neg_FG_cs ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm16Ma)

Call:
lm(formula = neg_FG_cs ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.176275 -0.073597  0.006977  0.066086  0.222399 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)  
(Intercept)      3.939e-01  1.596e+00   0.247   0.8062  
sqrt(INR)       -1.963e-01  1.236e-01  -1.588   0.1192  
Male1           -4.352e-01  2.010e-01  -2.165   0.0357 *
T1_Age          -9.400e-03  5.274e-02  -0.178   0.8593  
White           -2.424e-02  2.817e-02  -0.860   0.3941  
T1_ICV          -8.224e-05  1.043e-03  -0.079   0.9375  
sqrt(INR):Male1  4.144e-01  1.904e-01   2.177   0.0347 *
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.09531 on 46 degrees of freedom
Multiple R-squared:  0.1066,    Adjusted R-squared:  -0.009973 
F-statistic: 0.9144 on 6 and 46 DF,  p-value: 0.4932

std_beta(lm16Ma)

Longitudinal

outlierTest(lm17F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm17Ma <- lm(pos_lLG_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm17Ma)

Call:
lm(formula = pos_lLG_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.178283 -0.047509 -0.008937  0.029333  0.177946 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)   
(Intercept)      3.171473   1.926671   1.646  0.11093   
sqrt(INR)        0.442784   0.134101   3.302  0.00263 **
Male1            0.726124   0.256473   2.831  0.00849 **
T1_Age          -0.083784   0.057364  -1.461  0.15527   
White            0.012479   0.032578   0.383  0.70457   
T1_ICV          -0.002017   0.001312  -1.537  0.13558   
sqrt(INR):Male1 -0.660317   0.240825  -2.742  0.01052 * 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.09086 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.4574,    Adjusted R-squared:  0.3411 
F-statistic: 3.934 on 6 and 28 DF,  p-value: 0.005631

std_beta(lm17Ma)

outlierTest(lm18F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm18Ma <- lm(pos_lITG_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm18Ma)

Call:
lm(formula = pos_lITG_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.18613 -0.07960 -0.01354  0.07937  0.19463 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)     -1.1134525  2.4061682  -0.463 0.647120    
sqrt(INR)        0.7083473  0.1674754   4.230 0.000226 ***
Male1            0.9299320  0.3203019   2.903 0.007125 ** 
T1_Age          -0.0027093  0.0716409  -0.038 0.970101    
White            0.0321223  0.0406858   0.790 0.436443    
T1_ICV           0.0002795  0.0016391   0.171 0.865837    
sqrt(INR):Male1 -0.8790167  0.3007599  -2.923 0.006796 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1135 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.4209,    Adjusted R-squared:  0.2968 
F-statistic: 3.392 on 6 and 28 DF,  p-value: 0.01216

std_beta(lm18Ma)

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm19Ma <- lm(pos_rLG_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm19Ma)

Call:
lm(formula = pos_rLG_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.209777 -0.022577  0.008025  0.033420  0.130932 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)      0.6513438  1.4778256   0.441 0.662782    
sqrt(INR)        0.4051405  0.1028604   3.939 0.000495 ***
Male1            0.0746646  0.1967237   0.380 0.707150    
T1_Age          -0.0286914  0.0440005  -0.652 0.519674    
White            0.0446099  0.0249885   1.785 0.085064 .  
T1_ICV          -0.0005878  0.0010067  -0.584 0.563949    
sqrt(INR):Male1 -0.0847129  0.1847214  -0.459 0.650063    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.06969 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.5253,    Adjusted R-squared:  0.4236 
F-statistic: 5.164 on 6 and 28 DF,  p-value: 0.001102

std_beta(lm19Ma)

outlierTest(lm20F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm20Ma <- lm(pos_rSPL_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm20Ma)

Call:
lm(formula = pos_rSPL_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.17477 -0.07169  0.01102  0.05353  0.18477 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)   
(Intercept)     -2.484343   2.022558  -1.228  0.22955   
sqrt(INR)        0.462011   0.140775   3.282  0.00277 **
Male1            0.641758   0.269237   2.384  0.02416 * 
T1_Age          -0.009015   0.060219  -0.150  0.88208   
White           -0.011287   0.034199  -0.330  0.74384   
T1_ICV           0.001540   0.001378   1.118  0.27308   
sqrt(INR):Male1 -0.569646   0.252810  -2.253  0.03226 * 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.09538 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.3469,    Adjusted R-squared:  0.2069 
F-statistic: 2.479 on 6 and 28 DF,  p-value: 0.04751

std_beta(lm20Ma)

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm21Ma <- lm(neg_rPTR_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm21Ma)

Call:
lm(formula = neg_rPTR_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.19942 -0.01818 -0.00027  0.01722  0.12794 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)      1.4604586  1.3273173   1.100 0.280568    
sqrt(INR)       -0.3547662  0.0923846  -3.840 0.000644 ***
Male1           -0.6052391  0.1766885  -3.425 0.001913 ** 
T1_Age          -0.0492362  0.0395193  -1.246 0.223134    
White           -0.0004490  0.0224436  -0.020 0.984182    
T1_ICV          -0.0003762  0.0009042  -0.416 0.680566    
sqrt(INR):Male1  0.5601276  0.1659085   3.376 0.002172 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.06259 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.4011,    Adjusted R-squared:  0.2728 
F-statistic: 3.125 on 6 and 28 DF,  p-value: 0.01795

std_beta(lm21Ma)

outlierTest(lm22F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm22Ma <- lm(neg_lCerebel_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm22Ma)

Call:
lm(formula = neg_lCerebel_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.10335 -0.02336 -0.00587  0.02449  0.08326 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)      5.952e-01  1.052e+00   0.566  0.57593    
sqrt(INR)       -3.672e-01  7.320e-02  -5.016 2.65e-05 ***
Male1           -4.982e-01  1.400e-01  -3.559  0.00135 ** 
T1_Age          -1.214e-02  3.131e-02  -0.388  0.70115    
White           -3.447e-02  1.778e-02  -1.938  0.06275 .  
T1_ICV          -4.535e-05  7.164e-04  -0.063  0.94998    
sqrt(INR):Male1  4.634e-01  1.315e-01   3.525  0.00148 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.0496 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.518, Adjusted R-squared:  0.4147 
F-statistic: 5.015 on 6 and 28 DF,  p-value: 0.001331

std_beta(lm22Ma)

outlierTest(lm23F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm23Ma <- lm(neg_rSFG_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm23Ma)

Call:
lm(formula = neg_rSFG_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.262848 -0.045948  0.007162  0.048026  0.226662 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)  
(Intercept)     -0.9737408  2.2314265  -0.436   0.6659  
sqrt(INR)       -0.3907123  0.1553129  -2.516   0.0179 *
Male1           -0.5187947  0.2970408  -1.747   0.0917 .
T1_Age           0.0911934  0.0664381   1.373   0.1808  
White           -0.0316536  0.0377311  -0.839   0.4086  
T1_ICV           0.0002696  0.0015201   0.177   0.8605  
sqrt(INR):Male1  0.5007620  0.2789180   1.795   0.0834 .
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.1052 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.3116,    Adjusted R-squared:  0.164 
F-statistic: 2.112 on 6 and 28 DF,  p-value: 0.08357

std_beta(lm23Ma)

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm24Ma <- lm(neg_lSFG_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm24Ma)

Call:
lm(formula = neg_lSFG_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.15901 -0.04040  0.01431  0.05354  0.10485 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)      2.7951785  1.5365576   1.819  0.07960 .  
sqrt(INR)       -0.5155758  0.1069483  -4.821 4.53e-05 ***
Male1           -0.6778312  0.2045419  -3.314  0.00255 ** 
T1_Age           0.0529327  0.0457492   1.157  0.25704    
White           -0.0002376  0.0259816  -0.009  0.99277    
T1_ICV          -0.0021157  0.0010467  -2.021  0.05290 .  
sqrt(INR):Male1  0.6071040  0.1920626   3.161  0.00376 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.07246 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.5445,    Adjusted R-squared:  0.4469 
F-statistic: 5.579 on 6 and 28 DF,  p-value: 0.0006574

std_beta(lm24Ma)

outlierTest(lm25F)

No Studentized residuals with Bonferonni p < 0.05
Largest |rstudent|:

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm25Ma <- lm(neg_lSLF_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm25Ma)

Call:
lm(formula = neg_lSLF_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.106059 -0.016623  0.000047  0.032360  0.067213 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)      2.174e-01  9.249e-01   0.235 0.815858    
sqrt(INR)       -2.704e-01  6.438e-02  -4.200 0.000245 ***
Male1           -3.747e-01  1.231e-01  -3.043 0.005050 ** 
T1_Age          -1.746e-03  2.754e-02  -0.063 0.949888    
White           -2.153e-02  1.564e-02  -1.376 0.179627    
T1_ICV           7.871e-05  6.301e-04   0.125 0.901476    
sqrt(INR):Male1  3.653e-01  1.156e-01   3.160 0.003767 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.04362 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.4394,    Adjusted R-squared:  0.3193 
F-statistic: 3.658 on 6 and 28 DF,  p-value: 0.008302

std_beta(lm25Ma)

contrasts(vols$Male) = c(1, 0)
#SE male
lm26M <- lm(neg_rHipp_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols)
summary(lm26M)

Call:
lm(formula = neg_rHipp_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols)

Residuals:
      Min        1Q    Median        3Q       Max 
-0.187588 -0.028662 -0.004849  0.034577  0.137558 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)    
(Intercept)      1.662e-01  5.541e-01   0.300 0.764818    
sqrt(INR)       -2.395e-01  7.220e-02  -3.317 0.001269 ** 
Male1           -3.084e-01  9.393e-02  -3.284 0.001412 ** 
T1_Age           3.351e-03  6.421e-03   0.522 0.602955    
White           -1.613e-02  1.258e-02  -1.282 0.202856    
T1_ICV           3.696e-05  4.107e-04   0.090 0.928473    
sqrt(INR):Male1  3.187e-01  8.866e-02   3.595 0.000506 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.06213 on 100 degrees of freedom
  (40 observations deleted due to missingness)
Multiple R-squared:  0.1557,    Adjusted R-squared:  0.1051 
F-statistic: 3.075 on 6 and 100 DF,  p-value: 0.008368

std_beta(lm26M)

contrasts(vols_match$Male) = c(1, 0)
#SE male
lm26Ma <- lm(neg_rHipp_lg ~ sqrt(INR) * Male + T1_Age + White + T1_ICV, data = vols_match)
summary(lm26Ma)

Call:
lm(formula = neg_rHipp_lg ~ sqrt(INR) * Male + T1_Age + White + 
    T1_ICV, data = vols_match)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.12769 -0.02291  0.00175  0.02237  0.13387 

Coefficients:
                  Estimate Std. Error t value Pr(>|t|)  
(Intercept)     -1.0169430  1.2465803  -0.816   0.4215  
sqrt(INR)       -0.2214022  0.0867651  -2.552   0.0165 *
Male1           -0.3683556  0.1659410  -2.220   0.0347 *
T1_Age           0.0061051  0.0371155   0.164   0.8705  
White           -0.0253821  0.0210784  -1.204   0.2386  
T1_ICV           0.0008779  0.0008492   1.034   0.3101  
sqrt(INR):Male1  0.3932683  0.1558168   2.524   0.0176 *
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 0.05878 on 28 degrees of freedom
  (18 observations deleted due to missingness)
Multiple R-squared:  0.3814,    Adjusted R-squared:  0.2488 
F-statistic: 2.877 on 6 and 28 DF,  p-value: 0.02597

std_beta(lm26Ma)

Residualize all volumes

full sample

residualize_models_cs <-
  vols %>% 
  gather(region, volume, pos_rSTG_cs:neg_lAG_cs) %>% 
  group_by(region) %>% 
  nest() %>% 
  mutate(
    residuals = map(
      data, 
      ~lm(volume ~ T1_ICV + White + T1_Age, data = .)$residuals
    )
  ) %>% 
  unnest() %>% 
  mutate(
    region = as.factor(region)
  )
residualize_models_lg <-
  vols %>% 
  filter(!is.na(pos_rSPL_lg)) %>% 
  gather(region, volume, pos_rSPL_lg:neg_rPTR_lg) %>% 
  group_by(region) %>% 
  nest() %>% 
  mutate(
    residuals = map(
      data, 
      ~lm(volume ~ T1_ICV + White + T1_Age + Interval, data = .)$residuals
    )
  ) %>% 
  unnest() %>% 
  mutate(
    region = as.factor(region)
  )

age matched

residualize_models_cs_agematch <-
  vols_match %>% 
  gather(region, volume, pos_rSTG_cs:neg_lAG_cs) %>% 
  group_by(region) %>% 
  nest() %>% 
  mutate(
    residuals = map(
      data, 
      ~lm(volume ~ T1_ICV + White + T1_Age, data = .)$residuals
    )
  ) %>% 
  unnest() %>% 
  mutate(
    region = as.factor(region)
  )
residualize_models_lg_agematch <-
  vols_match %>% 
  filter(!is.na(pos_rSPL_lg)) %>% 
  gather(region, volume, pos_rSPL_lg:neg_rPTR_lg) %>% 
  group_by(region) %>% 
  nest() %>% 
  mutate(
    residuals = map(
      data, 
      ~lm(volume ~ T1_ICV + White + T1_Age + Interval, data = .)$residuals
    )
  ) %>% 
  unnest() %>% 
  mutate(
    region = as.factor(region)
  )

Visualize

Cross-sectional

residualize_models_cs <-
  residualize_models_cs %>% 
  mutate(
    region_named = factor(
      region,
      levels = c(
        "pos_rCV_cs", 
        "pos_rMB_cs", 
        "pos_rPTR_cs", 
        "pos_rSFG_cs_motor", 
        "pos_rPCG",
        "pos_rSFG_cs_dlPFC",
        "pos_rMFG_cs",
        "pos_lLOC_cs",
        "pos_lIFG_cs",
        "pos_rSTG_cs",
        "neg_lAG_cs",
        "neg_rThal_cs", 
        "neg_lHipp_cs", 
        "neg_rHipp_cs", 
        "neg_lThal_cs", 
        "neg_FG_cs"
      ),
      labels = c(
        "Right CV (3, -47, -35)",
        "Right midbrain (3, -32, -4)",
        "Right PTR (40, -55, 2)",
        "Right SFG (9, -8, 54)",
        "Right PoG (15, -32, 69)",
        "Right SFG (11, 46, 45)",
        "Right MFG (51, 38, 5)",
        "Left LOG (-23, -89, 3)",
        "Left IFG (-48, 46, -8)",
        "Right STG (63, -47, 19)",
        "Left AG (-46, -39, 33)",
        "Right thalamus (13, -23, 8)",
        "Left CGH (-29, -19, -33)",
        "Right CGH (23, -14, -23)",
        "Left thalamus (-9, -24, 10)",
        "Left FG (-49, -49, -16)"
      )
    ),
    Sex = factor(
      Male, 
      levels = c(0, 1),
      labels = c("Female", "Male")
    )
  )

glimpse(residualize_models_cs_agematch)

Observations: 848
Variables: 24
$ region          <fct> pos_rSTG_cs, pos_rSTG_cs, pos_rSTG_cs, pos_rSTG_cs, pos_rSTG_cs, pos_rSTG…
$ residuals       <dbl> -0.029977783, 0.179923920, 0.314162377, 0.035453172, -0.149856670, 0.0996…
$ ID              <dbl> 18, 22, 23, 25, 32, 35, 39, 42, 43, 46, 48, 49, 54, 60, 64, 65, 75, 76, 7…
$ T1_ICV          <dbl> 1352.300, 1317.275, 1359.031, 1347.595, 1327.863, 1339.837, 1338.875, 134…
$ White           <dbl> 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 1, 1, 0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 0…
$ T1_Age          <dbl> 11.09, 11.22, 11.94, 11.67, 11.53, 11.74, 11.22, 11.95, 11.70, 11.65, 11.…
$ Sex             <fct> Female, Male, Female, Male, Female, Male, Female, Male, Female, Male, Mal…
$ INR             <dbl> 1.2144654, 1.1273735, 1.0160947, 0.7463322, 1.0160947, 1.4021141, 1.27994…
$ INR_Sex         <dbl> 1.2144654, 2.2547470, 1.0160947, 1.4926645, 1.0160947, 2.8042282, 1.27994…
$ pos_rSPL_lg     <dbl> NA, 0.007105, 0.014721, -0.006227, NA, 0.074150, NA, -0.085309, 0.076353,…
$ pos_rLG_lg      <dbl> NA, 0.011626, -0.095641, -0.064236, NA, 0.044319, NA, 0.021908, -0.063848…
$ pos_lITG_lg     <dbl> NA, 0.096546, 0.092174, -0.192735, NA, -0.061005, NA, 0.274067, -0.080808…
$ pos_lLG_lg      <dbl> NA, 0.027371, -0.100991, 0.026319, NA, 0.135742, NA, -0.001221, -0.140797…
$ neg_rHipp_lg    <dbl> NA, -0.180199, 0.011528, 0.066754, NA, -0.023231, NA, 0.054839, 0.072891,…
$ neg_lSLF_lg     <dbl> NA, -0.025449, 0.029119, 0.103692, NA, -0.113493, NA, -0.045815, 0.046227…
$ neg_lSFG_lg     <dbl> NA, 0.121833, -0.134355, 0.221174, NA, -0.026029, NA, -0.067296, 0.053502…
$ neg_rSFG_lg     <dbl> NA, 0.031358, 0.300343, 0.073874, NA, 0.022223, NA, 0.059198, 0.063420, 0…
$ neg_lCerebel_lg <dbl> NA, -0.041313, -0.006640, 0.018558, NA, -0.146236, NA, -0.078917, 0.02951…
$ neg_rPTR_lg     <dbl> NA, 0.031497, -0.236286, -0.006042, NA, -0.042279, NA, -0.008907, 0.02144…
$ Interval        <dbl> NA, 1.67, 1.88, 1.95, NA, 2.58, NA, 1.82, 1.78, 1.22, 2.49, 1.77, 1.80, 1…
$ Male            <fct> 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 1, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0…
$ age_matched     <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1…
$ volume          <dbl> -0.076392, 0.022881, 0.095106, -0.124102, -0.271639, -0.069962, 0.057077,…
$ region_named    <fct> "Right STG (63, -47, 19)", "Right STG (63, -47, 19)", "Right STG (63, -47…

residualize_models_cs %>% 
  filter(str_detect(region, "neg")) %>% 
  ggplot(aes(INR, residuals, color = Sex, fill = Sex)) +
  geom_point(size = 3, alpha = 1/2) +
  geom_smooth(method = "lm", formula = y ~ sqrt(x), size = 2) +
  scale_x_continuous(breaks = seq.int(0, 2, .25)) +
  scale_color_manual(
    values = c("darkred", "royalblue4")
  ) +
  scale_fill_manual(
    values = c("darkred", "royalblue4")
  ) +
  theme_minimal() +
  theme(
    legend.title = element_blank(),
    axis.title = element_text(size = 24),
    axis.text = element_text(size = 18),
    legend.text = element_text(size = 22),
    strip.text = element_text(size = 18),
    legend.key.size = unit(2, 'lines')
  ) +
  facet_wrap(.~region_named) +
  labs(
    y = "Volume in early adolescence\n(residuals)",
    x = "Family income-to-needs ratio"
  )
ggsave(
  "~/Desktop/ELS/income_TBM/income_TBM_sync/plots/neg_cross-sectional.png",
  height = 11,
  width = 13
)

residualize_models_cs_agematch %>% 
  filter(str_detect(region, "neg")) %>% 
  ggplot(aes(INR, residuals, color = Sex, fill = Sex)) +
  geom_point(size = 3, alpha = 1/2) +
  geom_smooth(method = "lm", formula = y ~ sqrt(x), size = 2) +
  scale_x_continuous(breaks = seq.int(0, 2, .25)) +
  scale_color_manual(
    values = c("darkred", "royalblue4")
  ) +
  scale_fill_manual(
    values = c("darkred", "royalblue4")
  ) +
  theme_minimal() +
  theme(
    legend.title = element_blank(),
    axis.title = element_text(size = 24),
    axis.text = element_text(size = 18),
    legend.text = element_text(size = 22),
    strip.text = element_text(size = 18),
    legend.key.size = unit(2, 'lines')
  ) +
  facet_wrap(.~region_named) +
  labs(
    y = "Volume in early adolescence\n(residuals)",
    x = "Family income-to-needs ratio"
  )
ggsave(
  "~/Desktop/ELS/income_TBM/income_TBM_sync/plots/neg_cross-sectional_agematch.png",
  height = 11,
  width = 13
)

residualize_models_cs %>% 
  filter(str_detect(region, "pos")) %>% 
  ggplot(aes(INR, residuals, color = Sex, fill = Sex)) +
  geom_point(size = 3, alpha = 1/2) +
  geom_smooth(method = "lm", formula = y ~ sqrt(x), size = 2) +
  #scale_x_continuous(breaks = seq.int(0, 2, .2)) +
  scale_color_manual(
    values = c("darkred", "royalblue4")
  ) +
  scale_fill_manual(
    values = c("darkred", "royalblue4")
  ) +
  theme_minimal() +
  theme(
    legend.title = element_blank(),
    axis.title = element_text(size = 24),
    axis.text = element_text(size = 18),
    legend.text = element_text(size = 22),
    strip.text = element_text(size = 18),
    legend.key.size = unit(2, 'lines')
  ) +
  facet_wrap(.~region_named) +
  labs(
    y = "Volume in early adolescence\n(residuals)",
    x = "Family income-to-needs ratio"
  )
ggsave(
  "~/Desktop/ELS/income_TBM/income_TBM_sync/plots/pos_cross-sectional.png",
  height = 12,
  width = 15.5
)

residualize_models_cs_agematch %>% 
  filter(str_detect(region, "pos")) %>% 
  ggplot(aes(INR, residuals, color = Sex, fill = Sex)) +
  geom_point(size = 3, alpha = 1/2) +
  geom_smooth(method = "lm", formula = y ~ sqrt(x), size = 2) +
  #scale_x_continuous(breaks = seq.int(0, 2, .2)) +
  scale_color_manual(
    values = c("darkred", "royalblue4")
  ) +
  scale_fill_manual(
    values = c("darkred", "royalblue4")
  ) +
  theme_minimal() +
  theme(
    legend.title = element_blank(),
    axis.title = element_text(size = 24),
    axis.text = element_text(size = 18),
    legend.text = element_text(size = 22),
    strip.text = element_text(size = 16),
    legend.key.size = unit(2, 'lines')
  ) +
  facet_wrap(.~region_named) +
  labs(
    y = "Volume in early adolescence\n(residuals)",
    x = "Family income-to-needs ratio"
  )
ggsave(
  "~/Desktop/ELS/income_TBM/income_TBM_sync/plots/pos_cross-sectional_agematch.png",
  height = 12,
  width = 15.5
)

Longitudinal

residualize_models_lg <-
  residualize_models_lg %>% 
  mutate(
    region_named = factor(
      region,
      levels = c(
        "pos_lLG_lg",
        "pos_lITG_lg",
        "pos_rLG_lg",
        "pos_rSPL_lg",
        "neg_rPTR_lg",
        "neg_lCerebel_lg",
        "neg_rSFG_lg",
        "neg_lSFG_lg",
        "neg_lSLF_lg",
        "neg_rHipp_lg"
      ),
      labels = c(
        "Left LG (-20, -75, -5)",
        "Left ITG (-48, 1, -40)",
        "Right LG (12, -73, -9)",
        "R SPL (27, -60, 56)",
        "Right PTR (24, -70, 7)",
        "Left cerebellum (-18, -66, -21)",
        "Right SFG (17, -5, 58)",
        "Left SFG (-16, 62, -10)",
        "Left SLF (-33, -44, 16)",
        "Right CGH (26, -33, -7)"
      )
    ),
    Sex = factor(
      Male, 
      levels = c(0, 1),
      labels = c("Female", "Male")
    )
  ) 

residualize_models_lg_agematch <-
  residualize_models_lg_agematch %>% 
  mutate(
    region_named = factor(
      region,
      levels = c(
        "pos_lLG_lg",
        "pos_lITG_lg",
        "pos_rLG_lg",
        "pos_rSPL_lg",
        "neg_rPTR_lg",
        "neg_lCerebel_lg",
        "neg_rSFG_lg",
        "neg_lSFG_lg",
        "neg_lSLF_lg",
        "neg_rHipp_lg"
      ),
      labels = c(
        "Left LG (-20, -75, -5)",
        "Left ITG (-48, 1, -40)",
        "Right LG (12, -73, -9)",
        "R SPL (27, -60, 56)",
        "Right PTR (24, -70, 7)",
        "Left cerebellum (-18, -66, -21)",
        "Right SFG (17, -5, 58)",
        "Left SFG (-16, 62, -10)",
        "Left SLF (-33, -44, 16)",
        "Right CGH (26, -33, -7)"
      )
    ),
    Sex = factor(
      Male, 
      levels = c(0, 1),
      labels = c("Female", "Male")
    )
  ) 

residualize_models_lg %>% 
  filter(str_detect(region, "neg")) %>% 
  ggplot(aes(INR, residuals, color = Sex, fill = Sex)) +
  geom_point(size = 3, alpha = 1/2) +
  geom_smooth(method = "lm", formula = y ~ sqrt(x), size = 2) +
  scale_x_continuous(breaks = seq.int(0, 2, .25)) +
  scale_color_manual(
    values = c("darkred", "royalblue4")
  ) +
  scale_fill_manual(
    values = c("darkred", "royalblue4")
  ) +
  theme_minimal() +
  theme(
    legend.title = element_blank(),
    axis.title = element_text(size = 24),
    axis.text = element_text(size = 18),
    legend.text = element_text(size = 22),
    strip.text = element_text(size = 18),
    legend.key.size = unit(2, 'lines')
  ) +
  facet_wrap(.~region_named) +
  labs(
    y = "Change in volume from earlier to later adolescence\n(residuals)",
    x = "Family income-to-needs ratio"
  )
ggsave(
  "~/Desktop/ELS/income_TBM/income_TBM_sync/plots/neg_longitudinal.png",
  height = 11,
  width = 13
)

residualize_models_lg_agematch %>% 
  filter(str_detect(region, "neg")) %>% 
  ggplot(aes(INR, residuals, color = Sex, fill = Sex)) +
  geom_point(size = 3, alpha = 1/2) +
  geom_smooth(method = "lm", formula = y ~ sqrt(x), size = 2) +
  scale_x_continuous(breaks = seq.int(0, 2, .25)) +
  scale_color_manual(
    values = c("darkred", "royalblue4")
  ) +
  scale_fill_manual(
    values = c("darkred", "royalblue4")
  ) +
  theme_minimal() +
  theme(
    legend.title = element_blank(),
    axis.title = element_text(size = 24),
    axis.text = element_text(size = 18),
    legend.text = element_text(size = 22),
    strip.text = element_text(size = 18),
    legend.key.size = unit(2, 'lines')
  ) +
  facet_wrap(.~region_named) +
  labs(
    y = "Change in volume from earlier to later adolescence\n(residuals)",
    x = "Family income-to-needs ratio"
  )
ggsave(
  "~/Desktop/ELS/income_TBM/income_TBM_sync/plots/neg_longitudinal_agematch.png",
  height = 11,
  width = 13
)

residualize_models_lg %>% 
  filter(str_detect(region, "pos")) %>% 
  ggplot(aes(INR, residuals, color = Sex, fill = Sex)) +
  geom_point(size = 3, alpha = 1/2) +
  geom_smooth(method = "lm", formula = y ~ sqrt(x), size = 2) +
  scale_x_continuous(breaks = seq.int(0, 2, .25)) +
  scale_color_manual(
    values = c("darkred", "royalblue4")
  ) +
  scale_fill_manual(
    values = c("darkred", "royalblue4")
  ) +
  theme_minimal() +
  theme(
    legend.title = element_blank(),
    axis.title = element_text(size = 24),
    axis.text = element_text(size = 18),
    legend.text = element_text(size = 22),
    strip.text = element_text(size = 18),
    legend.key.size = unit(2, 'lines')
  ) +
  facet_wrap(.~region_named) +
  labs(
    y = "Change in volume from earlier to later adolescence\n(residuals)",
    x = "Family income-to-needs ratio"
  )
ggsave(
  "~/Desktop/ELS/income_TBM/income_TBM_sync/plots/pos_longitudinal.png",
  height = 11,
  width = 13
)

residualize_models_lg_agematch %>% 
  filter(str_detect(region, "pos")) %>% 
  ggplot(aes(INR, residuals, color = Sex, fill = Sex)) +
  geom_point(size = 3, alpha = 1/2) +
  geom_smooth(method = "lm", formula = y ~ sqrt(x), size = 2) +
  scale_x_continuous(breaks = seq.int(0, 2, .25)) +
  scale_color_manual(
    values = c("darkred", "royalblue4")
  ) +
  scale_fill_manual(
    values = c("darkred", "royalblue4")
  ) +
  theme_minimal() +
  theme(
    legend.title = element_blank(),
    axis.title = element_text(size = 24),
    axis.text = element_text(size = 18),
    legend.text = element_text(size = 22),
    strip.text = element_text(size = 18),
    legend.key.size = unit(2, 'lines')
  ) +
  facet_wrap(.~region_named) +
  labs(
    y = "Change in volume from earlier to later adolescence\n(residuals)",
    x = "Family income-to-needs ratio"
  )
ggsave(
  "~/Desktop/ELS/income_TBM/income_TBM_sync/plots/pos_longitudinal_agematch.png",
  height = 11,
  width = 13
)